Ray Fortna and I were discussing potential applications of satellite photos for everyday users, as enabled by applications such as Google Earth.  It’s especially powerful to be able to see how the satellite photos change over time.  There are obvious applications for real estate agents, house shoppers, tourists, and so forth. 

 

Another equally compelling application is to give people a better understanding of the world and some of the bad things happening in it.  Satellites can go places where journalists can’t a give a better sense of scale and impact.  The following BBC photos were blogged in One Man’s Trash » from the salmon  showing the destruction caused by Zimbabwe’s slum “cleanup” project.

 

 

“One month ago, Zimbabwe’s brutal leader Robert Mugabe began Operation Murambatsvina, which literally translates to “Operation Drive Out Trash.” The “trash”? 250,000 to 1.5 million of Zimbabwe’s urban poor who have had their homes bulldozed or burnt down by Mugabe’s police. The newly homeless are left to sleep outside in the freezing winter nights or are being pushed into rural areas, where they face dangerous food shortages and likely starvation. “

Scott Haug is blogging the JavaOne conference.

Berry Reverse Phone Lookup is a Blackberry application that allows you to conveniently enter a phone number into your Blackberry and view the corresponding name and address. Handy when you need to figure out who called you or where someone lives. It’s more convenient than using the web browser because the input field is locked to numeric mode and limits entry to 10 digits

The following are rough (Blackberry jotted) notes from a panel discussion on classified and job metasearch at the Vertical LEAP Search Engine Conference.

There were a number of questions concerning the viability of job and classifieds metasearch as a business, the impact of metasearch on existing job and classified players like Monster and Craigs List, and the impact of new entrants such as Google.

Panel Participants

Moderator
John Zappe- classified intelligence report

Participants
Guatam Godhwani, simply hired
Craig Donato, Oodle
Konstantun Guericke, LinkedIn
Garrett Price, Kijiji/Ebay

Why are classifieds “hot” right now
GG: We are currently undergoing a paradigm shift– personalized aggregation of classified content. Rather than just being an advertising medium, new sites allow leads to be qualified and transactions to take place in the context of the ad.

CD: classified ads are hot because internet adoption on a local level has reached critical mass. Free changes everythung. How is free a business model? Most powerful when everything is free. Monetize further down the cycle– the qualified lead or the hire. Employers pay for real value.

The basic listing will be free– will charge us upgrades for print or virtual listings or performance. Become more of a partner with the advertisor.

GP- the community will tell us where we are adding value

How dependent are you on the willingness of unintentional partners to provide free content?
GG– choice for users has gone up. Hard for users to conduct comprehensive search. We send more traffic to each of these sites– we see you as a distribution channel.

How does social networking fit into classifieds?
KG– less applicable to classifieds for products. Real estate agents, could be relevant. High value transactions– choosing auditing, law firm, hiring. How many of you have gotten a job because someone referred you?

CD– the set of listings out there will grow more not less fragmented. Self published content and communities.

Buzz is that Google will enter a classifieds market. What impact will this have on metasearch enginges.
GG- Google will validate the market. Google is less successfuly in specialized search. Lots of specialized work required for jobs, community, etc.

Will metasearch lead to a decay in Monster’s position?
Monster charges 395 for a job. Craigs list charges 75 or free. When you aggregate all jobs are equalized. Why won’t businesses shift to the lower cost options.

GG– some users will continue to go to destnation sites. Will Monster come under price pressures? Yes, that will happen regardless of what Simply Hired does. Monster will adapt to the new environment. We think they will embrace payment for results.

KG– at linked in we are both charging for listings and aggregating via simplyhired. Offer advantages for those who pay– a pool of people who don’t go to Monster, reference checks, see connections to applicants.

GG– the winning company in the market will have the best content and community that takes them where they need to go.

Have you considered using social networks to improve relevance?
KG– LinkedIn believes that social networks are most relevant in hiring decisions and the selection of trusted professional services (real estate agents, lawyers, accountants, etc.)

What keeps users coming back to a vertical search site?
GG: we will establish an ongoing relationship that keeps users coming back

Greasemonkey requires manual configuration for each user script that is installed. This is a hassle for a user who wants a consistent configuration across several machines, and untenable for an organization that wants consistent configurations across hundreds or thousands of machines.

These problems will grow as user scripts become more prevelant. It will become important to be able to upgrade script versions, to automatically add cool new (safe) scripts, and to block hostile scripts. Something like this is probably essentially if Greasemonkey is to gain any traction in the workplace.

One obvious solution is to push the configuration on to a centralized server and use the local filesystem merely as a cache.

It’s fairly easy to achieve much of this vision without changing Greasemonkey at all. The “uber user script” at the bottom of this post has no hardwired logic of its own, but uses XmlHttpRequest to pass the current URL to a web server that can decide what logic to run in response to that web page.

You can imagine backing with a variety of different web based configuration tools optiimized for individuals, corporations, and groups. It would also be possible to make the existing Greasemonkey dialogs talk to the web server rather than the filesystem to record user changes.

This isn’t just a tool for IT Nazis. Imagine being able to have communities of people sharing script recommendations.

To make this more efficient, we would want to respect the http cache headers on the fetched script so that it doesn’t have to be fetched repeatedly for the same URL. We would also want a way for Greasemonkey to ask the server for all of the URL patterns that the server might be interested in, so that on other URLs the server doesn’t need to be hit at all.

// uberuserscript
//
// ==UserScript==
// @name Uber User Script
// @namespace http://www.example.com
// @description Allow a server to specify greasemonkey user script configuration
// @include *
// ==/UserScript==
GM_xmlhttpRequest
({
method:’GET’,
url: ‘http://greasemonkeyconfig.com/userscript?url=’ + escape(document.location),
onload:function(results) {
eval(results.responseText)
}
});

Google is indeed smart. They have released a Google Maps API that allows you to embed Google maps in any web page.

Notice the bold point below.

“The Maps API is a free beta service, available for any web site that is free to consumers. Google retains the right to put advertising on the map in the future. Please see the terms of use for more information.”

Google gets the benefit of an army of web developers creating value added applications on top of Google maps, like Busmonster, which collectively become a distributed canvas for Google’s advertisements.

I’m back from the Gnomedex 5 conference.

The headlines are increasing support from mainstream companies for RSS, including platform support that makes it write RSS enabled applications and extensions to RSS itself that increase the scope of content and media that work well in RSS.

In this post I’d like to consider specifically the extensions to RSS.

Dean Hachamovitch and the Microsoft folks had a proposed RSS extension called Simple List Extensions, licensed under the creative commons license. Simple List Extensions should make things like RSS job feeds, to do lists, and other list based subscriptions work much better over RSS.

The problem is that RSS today is optimized for feeds of individuals posts that accumulate over time. It doesn’t work well for lists whose items can reordered, updated, and deleted– these changes will show up as entirely new posts or not all. And there is now way for a feed to specify additional sorts and filter that would be relevant to the content in that feed.

The simple list extensions allow a RSS feed to be designated as a list, with metadata to describe the relevant sort orders and category filters that might applied. Downlevel clients will ignore the extensions, smart clients will present a richer and better UI based on them.

Amazon had created an RSS feed of their wish lists with these extensions, and Dean demonstrated an aggregator taking advantage of them. He could click on a “books” link to filter the feed items to just books and then “price” to sort by price. The sorting and filterign are very fast because they happened entirely on the client.

The snippet of XML for doing this looks something like the following:

<cf:listinfo>
<aws:DateAdded type=”text”>Date Added <aws:Price type=”number”>Price</aws:Price>
…
</cf:listinfo>

Simple List Extensions points the way towards what is likely to be a number of semantic hints to RSS clients. I don’t worry to much about Simple List Extensions adoption– it solves a very simple problem simply, and has a reasonable Creative Commons license– but I foresee the potential for ferocious and dueling standards over future extensions.

The one down side in all of this is that support for Simple List Extensions seems to be tied up in the slow moving Longhorn bandwagon. If Microsoft is serious about evangelizing the standard they have support for it now, not when Longhorm ships. They don’t even need to ship client bits to do this. If their portal consumes RSS they could extend the portal to understand the list extensions and evangelize them among content author.

If Microsoft doesn’t do this, other companies will. I can easily imagine Yahoo and Google collaborating to define a competing standard for RSS lists and implementing it on their portals, which already support RSS. From the perspective of someone who would like to see the functionality arrive sooner rather than later and achieve wide adoption, a standard that was born in this fashion might be preferable.

Relevant Links:

• Simple List Extensions spec spec
• Dean’s Gnomedex blog entry
• MSDN Longhorn RSS page,

I will be attending Gnomedex 5.0.

If you’re planning on attending and interested in chatting in person, please get in touch with me via email.

Nokia - Nokia develops a new browser for Series 60 by using open source software

Thanks to Scott Haug for the tip off.

Here are some recent additions to the small but growing roster of Berry 411 plugins:

Amazon: Search Amazon products
mlb: Show today’s baseball games and scores
rphone: Do a reverse phone lookup from the given phone address to the corresponding name and number